Handheld meter robustness testing

[floobydust]

Maybe Joe could set up his own agency and standards lol even for pre-screening.

CSA charges out CAD $425/hour and UL very good quality but top price, even more expensive. This is why people are going with Intertek and other agencies. They really have a monopoly - their own "university/training", expert knowledge system and free access to all standards, senior certifiers etc. They also have quotas and are pushed to make money, pound out the files every month.

Because of this, I shamefully admit to getting chinese assessment and certificate on one product simply because they are fast, efficient, low cost, high volume certs there it's major manufacturing, and if you catch any errors/omissions the quality is fine.
UL/CSA can be terribly slow and expensive, you'll get a junior certifier that knows nothing and then gets hung up on on a small design issue and wastes time dollars. Then the guy gets up to speed and leaves to another department. They'll give quotations but go way over it. No accountability for cost or the results, if in the end they bungle a certification it's entirely the manufacturer's problem/cost, per the signed contract.

[...] One could look at the input current to a multimeter during an ohms-function overload, to see the PTC heat up and settle. [...]

I meant long-term overload, such as mains applied when on the ohms function. PTC holding current might be 10mA? hard to tell temperature in the heatshrink tubing sleeve but it's going be hot.

[Fungus]

....generally multimeters have a separate 10MOhm voltage input and a 10kOhm+PTC resistance/continuity input.

Keep those stories coming.  We're here all night folks.

I found another meter that has a separate path for voltage that doesn't go via the surge resistor/PTC.

It's none other than our little friend the Fluke 101!



The voltage input goes through R2/R3/R4/R5/R6/R7/R8. The Ohms/diode/continuity/etc. goes the other way.

See, I'm not crazy! 

[Neutrion]

Especialy the low energy spikes will be hard to find any info about, because they don't cause imediate failure so it would be difficult to asess what would be the most common but already harmful energy and voltage level.But I think something well within Joes generators range.

You're going to have to define and quantify 'low energy' and fully specify the circuit characteristics and other test conditions for any further discussion to have meaning.  I wouldn't call the jqs^TM transients 'low energy'.

That is what I meant. There must be some statistical data about the occurrance of these. Possibly the weaker the more common. So there must be a minimum value which definitely causes some damage.
And yes I think also  that 20 Joule is possibly well beyond that level, but also well below the rating of any surge arrestor. So although zapping an arrestor with it would not mean that that arrestor would also perform well on its rated level, but if it fails and lets through too high voltages or clamps too slow than that is definitely a fail.

[bdunham7]

Here's two of them side by side:

So you have your choice of a CAT III/1000V meter with 600V fuses and quite a population of PTCs and MOVs or a CAT III/1000V meter with even less appropriate fuses and no MOVs.  Let's call the versions 'fail' and 'didn't even try'.

[joeqsmith]

....generally multimeters have a separate 10MOhm voltage input and a 10kOhm+PTC resistance/continuity input.

Keep those stories coming.  We're here all night folks.

I found another meter that has a separate path for voltage that doesn't go via the surge resistor/PTC.

It's none other than our little friend the Fluke 101!
..

The voltage input goes through R2/R3/R4/R5/R6/R7/R8. The Ohms/diode/continuity/etc. goes the other way.

See, I'm not crazy! 

We need to start at the beginning where you wrote:

I haven't looked closely at this particular Uni-T but generally multimeters have a separate 10MOhm voltage input and a 10kOhm+PTC resistance/continuity input.

Take your Fluke 101 and lift one of the follow pins to remove the PTC as you suggest (it's not needed for voltage).   Next apply both a DC and AC voltage to the input of the meter.   Keep the signal at a safe level (<10V).    Leave the common connector attached and remove the other.  Next inject the test voltage to the PCB where you lifted the pin and measure both the AC and DC voltages.   Report your findings. 

[joeqsmith]

Here's two of them side by side:

So you have your choice of a CAT III/1000V meter with 600V fuses and quite a population of PTCs and MOVs or a CAT III/1000V meter with even less appropriate fuses and no MOVs.  Let's call the versions 'fail' and 'didn't even try'.

And lets not forget the third version that has even less parts.   These two are in a much nicer class. 

[EEVblog]

....generally multimeters have a separate 10MOhm voltage input and a 10kOhm+PTC resistance/continuity input.

Keep those stories coming.  We're here all night folks.

I found another meter that has a separate path for voltage that doesn't go via the surge resistor/PTC.

It's none other than our little friend the Fluke 101!



The voltage input goes through R2/R3/R4/R5/R6/R7/R8. The Ohms/diode/continuity/etc. goes the other way.

See, I'm not crazy! 

(WTB I have not been following any argument in this thread, so I don't know the history here)
The surge resistor and PTC are still used in the voltage range to clamp via the two series MOVs to ground.
Once you have a clamp like for any main energy spike, then you don't need any major protection on the voltage input because of the attenuator. Maybe just a basic extra transistor clamp or something just in case.

[EEVblog]

...
But then you have companies like Uni-T that have a reputation for changing and omitting parts in models on a whim.
...

That's anecdotal, and even though it matters for a company's reputation or "vibes", objectively you would have to comparatively test to see how it affects the product's performance or safety.

Fungus posted this before, two UT-61E's.
I don't need a compartive test to tell me that one lacks all the MOV's and has smaller and lower rated fuses than the other.
EDIT: Here is my UT61E added to the right side:



And there are other examples of Uni-T meters doing this. I've even had it myself were a meter I have differs a lot from one someone else has.

[joeqsmith]

...
The surge resistor and PTC are still used in the voltage range to clamp via the two series MOVs to ground.
...

No argument.  It was their comments like this:

The PTCs aren't on the "voltage measurement" part of a multimeter circuit, they're on the much lower-impedance "ohms/continuity/etc" part of the circuit.

why I suggested they have a fill in the blank degree.   One way to learn is by doing, which is what I suggested they try.   It's a simple test to remove that PTC and see what happens with the voltage measurements.  Sure we know the outcome, but it appears that fungus may learn something from it.   

[joeqsmith]

...
But then you have companies like Uni-T that have a reputation for changing and omitting parts in models on a whim.
...

That's anecdotal, and even though it matters for a company's reputation or "vibes", objectively you would have to comparatively test to see how it affects the product's performance or safety.

Fungus posted this before, two UT-61E's.
I don't need a compartive test to tell me that one lacks all the MOV's and has smaller and lower rated fuses than the other.
EDIT: Here is my UT61E added to the right side:



And there are other examples of Uni-T meters doing this. I've even had it myself were a meter I have differs a lot from one someone else has.

The discussion evolves around safety and certifications.  I don't know what the pass/fail criteria is for safety, only what I was told from two different companies that had very different interpretations.   I assume all three flavors have been certified for use in CAT III 600 or what ever they are marked for.  Is the cert worth anything?    Personally, I have bad vibes about the whole process.   

[Fungus]

(WTB I have not been following any argument in this thread, so I don't know the history here)
The surge resistor and PTC are still used in the voltage range to clamp via the two series MOVs to ground.
Once you have a clamp like for any main energy spike, then you don't need any major protection on the voltage input because of the attenuator. Maybe just a basic extra transistor clamp or something just in case.

A few pages back I said something about the volts-measurement path inside meters not going via the surge resistor+PTC and I got called out for it. The 101 is an example of a well regarded meter that does exactly that. 

The 10MOhms in that path will obviously attenuate a spike to the point where it can't damage the IC or it can be clamped by small components.

[Fungus]

EDIT: Here is my UT61E added to the right side:

That's three non-anecdotal variants!

Worse: You can see Uni-T deliberately planned to produce these unsafe variants after CAT certification by Intertek - the PCB has holes in it for different size fuse holders.

[EEVblog]

The 10MOhms in that path will obviously attenuate a spike to the point where it can't damage the IC or it can be clamped by small components.

Unless the resistors arc over. But there no way the input to the divider can go above the MOV clamping voltage, unless they fail, or you are talking about crazy small clamping times.
You design the 10M divider to easily handle the MOV clamping voltage and Bob's your uncle.

[EEVblog]

EDIT: Here is my UT61E added to the right side:

That's three non-anecdotal variants!
Worse: You can see Uni-T deliberately planned to produce these unsafe variants after CAT certification by Intertek - the PCB has holes in it for different size fuse holders.

To be fair, there is nothing wrong with that, provided that they are clearly marked and marketed as different model, but share the same PCB.
The fact that they are all labeled and sold as UT61E is troublesome.

[Fungus]

The surge resistor and PTC are still used in the voltage range to clamp via the two series MOVs to ground.

No argument.

What if the PTC heats up? Now the MOVs aren't doing much.

[EEVblog]

A few pages back I said something about the volts-measurement path inside meters not going via the surge resistor+PTC and I got called out for it. The 101 is an example of a well regarded meter that does exactly that. 

The Fluke 28 is another example. The 10M input resisotr is a ceramic jobbie, but in this case it does have an extra MOV on the other side of it.

[Fungus]

The Fluke 28 is another example.

 

My only point was that many meters have a path from the input jack to the IC that doesn't go via the PTC as shown in your input protection video.

If the PTC explodes during the first transient then your chances of surviving the second transient depend on the 10MOhm impedance in that path.

Edit: Yes, I know all about the surge resistor... all I'm saying is that the PTC could be out of action by eg. the user trying to measure mains AC on the Ohms range just before the lightning hits.

[bdunham7]

The Fluke 28 is another example. The 10M input resisotr is a ceramic jobbie, but in this case it does have an extra MOV on the other side of it.

These don't work quite the way you might think at first glance and you really need to look at schematics of meters (or at least I do) to see what really goes on.  I don't have a 28 schematic, but here's a what I think is a similar setup from an 87V/AN.  That first ceramic resistor Z2 is a 1M and is the sense resistor for ohms and some other stuff, the voltage and mV circuit actually does go through the PTC and surge resistor, the '10M' voltage divider resistor (Z1 in the diagram) is the one further away in your picture.

[joeqsmith]

The surge resistor and PTC are still used in the voltage range to clamp via the two series MOVs to ground.

No argument.

What if the PTC heats up? Now the MOVs aren't doing much.

Sure they do.  They continue to clamp the voltage and provide the majority of the return path.   Now if you select one of the functions that engages the low voltage clamps, then the MOVs are basically removed.     

My only point was that many meters have a path from the input jack to the IC that doesn't go via the PTC as shown in your input protection video.

Actually your original comments were quite clear:

The PTCs aren't on the "voltage measurement" part of a multimeter circuit, they're on the much lower-impedance "ohms/continuity/etc" part of the circuit.

I haven't looked closely at this particular Uni-T but generally multimeters have a separate 10MOhm voltage input and a 10kOhm+PTC resistance/continuity input.

You offered the Fluke 101 as an example and I suggested you actually test it by removing the PTC from the circuit and reporting your findings: 
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg3717439/#msg3717439

With you now back-pedaling on your original statements and suggesting your point was something completely different I am curious:

A) You removed the PTC and found the 101 could no longer read AC and/or DC voltages? 
B)  You removed the PTC and found the 101 could still read AC and/or DC voltages but you don't want me to eat too much crow?
C) You want to continue to believe your original statements and are concerned they are wrong?  (burying your head in the sand rather than confronting it)
D) You lack the soldering skills to lift one pin or the 101 costs too much and you are concerned about damaging it?
E) You're too lazy and can't be bothered to test your theory?

[AndrewBCN]

Which "UT61E" do you claim is certified? Which one do you think you'll get for 75 Euros? Better cross your fingers when you order...

The UT61E has been discontinued, so your entire argument is moot.

Oh, and just a note: as usual, you post pictures from others without attribution or mention of their origin. That picture of two different UT61E DMMs is by user Nisei and comes from this thread:
https://www.eevblog.com/forum/testgear/uni-t-ut61e-multimeter-teardown-photos/1000/

He clearly mentions that these are two different models of the UT61E.

[HKJ]

Meters often has 3 paths from the V input terminal:

1) The 10Mohm path, it do not really need a MOV/PTC protection, but often share the protection from 2)
2) A lower impedance path (Typical for mV/ohm/capacity/etc), on better meters it is protected with a MOV/PTC combination (Not all meters has this path, but most has).
3) Current output for ohm and capacity, it is always protected with a PTC and on better meters also with a MOV. The MOV may not do anything when the range is selected, but protect the rotary switch when not selected.

I have written some more about multimeter input protection here: https://lygte-info.dk/info/DMMDesignProtection%20UK.html

[AndrewBCN]

...
A) You removed the PTC and found the 101 could no longer read AC and/or DC voltages? 
B)  You removed the PTC and found the 101 could still read AC and/or DC voltages but you don't want me to eat too much crow?
C) You want to continue to believe your original statements and are concerned they are wrong? 
D) Lack the soldering skills to lift one pin or the 101 costs too much and you are concerned about damaging it?
E) You're too lazy and can't be bothered to test your theory?

Joe, you are assuming Fungus actually owns or has access to a Fluke 101. It's not that he is too lazy to test his theories, it's just that he can't. He really is the perfect armchair expert on DMMs that he has never even seen in person, much less taken apart.

[joeqsmith]

...
A) You removed the PTC and found the 101 could no longer read AC and/or DC voltages? 
B)  You removed the PTC and found the 101 could still read AC and/or DC voltages but you don't want me to eat too much crow?
C) You want to continue to believe your original statements and are concerned they are wrong? 
D) Lack the soldering skills to lift one pin or the 101 costs too much and you are concerned about damaging it?
E) You're too lazy and can't be bothered to test your theory?

Joe, you are assuming Fungus actually owns or has access to a Fluke 101. It's not that he is too lazy to test his theories, it's just that he can't. He really is the perfect armchair expert on DMMs that he has never even seen in person, much less taken apart.

You are assuming I am clueless on how to use a simple search engine.   Nice title change BTW.  Are we 12?   

How many Fluke meters do you own?  3 or so as I recall?

Five! (101, 187, 27FM, 37, 8060A)

Brymens? Only one! 

Referring to your recent comment:

The UT61E has been discontinued, so your entire argument is moot.

In France, does the word discontinued mean something is no longer available for purchase?   Or are you just trolling him? 

[rsjsouza]

...
But then you have companies like Uni-T that have a reputation for changing and omitting parts in models on a whim.
...

That's anecdotal, and even though it matters for a company's reputation or "vibes", objectively you would have to comparatively test to see how it affects the product's performance or safety.

Fungus posted this before, two UT-61E's.
I don't need a compartive test to tell me that one lacks all the MOV's and has smaller and lower rated fuses than the other.
EDIT: Here is my UT61E added to the right side:



And there are other examples of Uni-T meters doing this. I've even had it myself were a meter I have differs a lot from one someone else has.

The discussion evolves around safety and certifications.  I don't know what the pass/fail criteria is for safety, only what I was told from two different companies that had very different interpretations.   I assume all three flavors have been certified for use in CAT III 600 or what ever they are marked for.  Is the cert worth anything?    Personally, I have bad vibes about the whole process.   

In fairness, the two meters to the right were never subjected to testing as AndrewBCN pointed out. Also, Dave was an earlier production model.

Sure, this does not instill confidence but I have a BM857 that is quite different than the currently sold versions as well - my model is from 2002 and has no independent certification markings. It can happen to anyone, really.

[AndrewBCN]

...
But then you have companies like Uni-T that have a reputation for changing and omitting parts in models on a whim.
...

That's anecdotal, and even though it matters for a company's reputation or "vibes", objectively you would have to comparatively test to see how it affects the product's performance or safety.

Fungus posted this before, two UT-61E's.
I don't need a compartive test to tell me that one lacks all the MOV's and has smaller and lower rated fuses than the other.
EDIT: Here is my UT61E added to the right side:

...

And there are other examples of Uni-T meters doing this. I've even had it myself were a meter I have differs a lot from one someone else has.

I am not sure what your point is exactly. UNI-T is certainly not the only manufacturer that changes the PCB of a product during its lifecycle, or that populates the PCB with different components from one production batch to another.

Also the pictures of the discontinued UT61E that Fungus posted are from another EEVblog forum member Nisei, who duly noted (as can be seen on the PCB silk screen) that the two UT61E DMMs that he owns are different variants of the UT61E - the model on the left is a UT61E-GS and indeed it has better input protection and larger fuses.

Finally, what I was pointing out is that UNI-T again offers the UT61E+ in two different variants, the standard UT61E+ (which Joe is testing these days) and the Intertek-certified UT161E which costs only 2~3€ more. Whether they differ internally in terms of fuses and input protection is the question, and referring to 3~4 years old pictures of a different, discontinued model seems to me a bit irrelevant and a distraction.